Hi, I'm Anirudh!

Agricultural droughts can cause many serious hazards. Drought monitoring indices, namely Normalized Difference Vegetation Index (NDVI), Atmospherically Resistant Vegetation Index (ARVI), Soil Adjusted Vegetation Index (SAVI), and Enhanced Vegetation Index (EVI) have been used for an agricultural drought assessment. Satellite images from the Kolar region of Karnataka are used to calculate these indices. This paper proposes an integration model based on Convolutional Neural Networks (CNN) and a bio-inspired algorithm (Sparrow Search Algorithm (SSA) and Barnacles Mating Optimizer (BMO)) considering the indices as population. Performance is compared with the standalone CNN model in terms of efficiency. For the CNN, the accuracy, time taken for Epoch1, and time taken for Epoch2 is 91%, 16s (3s/step), and 2s (2s/step), respectively. For the CNN integrated with SSA, it is 94%, 3s (3s/step) and 0s (43ms/step), respectively. For the CNN integrated with BMO, it is 94%, 3s (2s/step) and 0s (46ms/step) respectively.

K-Nearest Neighbors (KNN), a simple and widely used algorithm, is extremely valuable in the field of machine learning models. Finding an optimal value of the nearest neighbor parameter in the KNN algorithm has been a problem that has existed for quite some time. In recent years, nature inspired algorithms have been widely emerging, with scope of being utilized in various engineering problems. This work proposes integrating a swarm intelligence optimization technique, the Sparrow Search Algorithm (SSA), for optimizing the process of finding K. The proposed hybrid model has been tested against four datasets of varying sizes and the results depict improvement in efficiency. The improvement is much larger as the size of the dataset increases, with around a 99% improvement in efficiency for the largest dataset tested on. KNN is also used to preprocess two of the datasets which have missing values, which result in improved accuracies of prediction.

Many meta heuristic optimization techniques have been developed to address exceedingly difficult optimization issues in numerous disciplines. Fireworks Algorithm (FWA) was proposed as a Swarm Intelligence Optimization technique, mimicking fire work explosions. This paper proposes using FWA in the field of Agricultural Drought Assessment. Droughts are complex natural disasters that need efficient mitigation techniques. Latest research in existing drought assessment tools suggest utilizing Convolutional Neural Network (CNN) models, often hybridized with other optimization strategies. Considering the four vegetation indices Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Atmospherically Resistant Vegetation Index (ARVI) and Soil Adjusted Vegetation Index (SAVI), a hybridized model that integrates FWA into CNN is proposed. This enhanced model can give up to 94% accuracy with very minimal loss for prediction of future drought. This model improves efficiency significantly.

Drought globally impacts many sectors, importantly agriculture and ecosystems. Drought impact mitigation is an immediate need, which can be achieved through reliable drought monitoring/prediction and early warning. Analyzing past patterns and predicting future drought conditions can significantly reduce the impact of drought on individuals, communities, and the environment. In this paper, the intelligent hybrid model for drought assessment is proposed. The proposed hybrid model integrates Convolutional Neural Network (CNN) coupled with a Sparrow Search Algorithm (SSA) for observing improvement in prediction efficiency. Various indices such as Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI), Enhanced Vegetation Index (EVI) and Atmospherically Resistant Vegetation Index (ARVI) are calculated for the drought assessment model using the satellite data from the Kolar region of Karnataka, India. The accuracy for the standalone CNN model is 91% whereas the accuracy of the CNN when coupled with SSA improved to 94%. The time taken to run the hybrid model also improved immensely. The model is also used to classify the predicted severity of drought into low, medium and severe based on the NDVI values computed.

Natural and complex climate disasters like drought have a number of underlying causes that are observed over timescales ranging from months to years. Sustaining natural resources for farming necessitate drought management plans wherein drought prediction is becoming powerful and flexible with intelligent techniques. It has been proved that machine learning and deep learning techniques are successful for drought prediction. Usage of ensemble hybrid intelligent learning algorithm is available for groundwater and gully erosion modeling but rarely emphasized for drought prediction in the literature. This paper discusses ensemble of Convolutional Neural Network (CNN) and Barnacles Mating Optimizer (BMO) to enhance the efficiency of a CNN model for drought prediction. The input for the proposed ensemble learning model includes the indices Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI), Atmospherically Resistant Vegetation Index (ARVI) and Enhanced Vegetation Index (EVI), which are calculated from satellite data taken for Kolar regions of Karnataka. predicted drought is classified into low drought, moderate drought and severe drought, based on the NDVI value. Improved results are observed.

Use of NoSQL to maintain databases has gained popularity in recent times. However, traditional relational databases provide structured data which helps in quick response time to execute certain queries. Currently there exist many algorithms that convert relational to unstructured databases, but vice versa is not available extensively. In this paper we propose an algorithm to convert NoSQL databases to relational databases, i.e., conversion of MongoDB databases to MySQL databases. Five datasets taken from GitHub, namely mobile accessories, bookstore, university database, restaurants and ecommerce store, are used as case studies to show the performance when stored as MongoDB database and as MySQL database. The input is given in JSON format and the output is obtained as MySQL tables. The execution time of the proposed algorithm and the time taken to execute queries are tabulated. It is observed that the execution time of the group-by queries improves by 75.33% when the database is converted to SQL.